Macro2Micro: A Rapid and Precise Cross-modal Magnetic Resonance Imaging Synthesis using Multi-scale Structural Brain Similarity
Sooyoung Kim, Joonwoo Kwon, Junbeom Kwon, Jungyoun Janice Min, Sangyoon Bae, Yuewei Lin, Shinjae Yoo, Jiook Cha
TL;DR
Macro2Micro addresses the challenge of deriving microstructural information from macrostructural MRI by introducing a GAN-based framework that encodes macro- and micro-scale brain information into two frequency branches via Octave Convolutions and enables their exchange during synthesis. The model employs a brain-focused patch discriminator and a perceptual loss to suppress artifacts and preserve individual biology, achieving notable improvements (e.g., about a 6.8% increase in SSIM over prior methods) in translating T1-weighted sMRI to FA and tractography. It demonstrates state-of-the-art performance across FA and tractography generation, preserves biologically relevant signals in downstream tasks, and offers ultra-fast inference, suggesting strong potential for real-time, multi-modal MRI rendering in clinical and research settings. While central-slice training limits peripheral accuracy, the approach establishes a scalable cross-modal, multi-scale mapping framework with significant implications for reducing MRI acquisition time and expanding multi-modal analyses.
Abstract
The human brain is a complex system requiring both macroscopic and microscopic components for comprehensive understanding. However, mapping nonlinear relationships between these scales remains challenging due to technical limitations and the high cost of multimodal Magnetic Resonance Imaging (MRI) acquisition. To address this, we introduce Macro2Micro, a deep learning framework that predicts brain microstructure from macrostructure using a Generative Adversarial Network (GAN). Based on the hypothesis that microscale structural information can be inferred from macroscale structures, Macro2Micro explicitly encodes multiscale brain information into distinct processing branches. To enhance artifact elimination and output quality, we propose a simple yet effective auxiliary discriminator and learning objective. Extensive experiments demonstrated that Macro2Micro faithfully translates T1-weighted MRIs into corresponding Fractional Anisotropy (FA) images, achieving a 6.8\% improvement in the Structural Similarity Index Measure (SSIM) compared to previous methods, while retaining the individual biological characteristics of the brain. With an inference time of less than 0.01 seconds per MR modality translation, Macro2Micro introduces the potential for real-time multimodal rendering in medical and research applications. The code will be made available upon acceptance.